In this paper we propose an effective and enhanced data preprocessing methodology which produces an efficient usage patterns and reduces the size of weblog down to 75-80% of its initial size. The experimental results are also shown in the following chapters. | ISSN:2249-5789 M Praveen Kumar et al, International Journal of Computer Science & Communication Networks,Vol 2(1), 55-60 An Effective Analysis of Weblog Files to improve Website Performance 1 , 2 Kumar, Babu, Rahaman, Reddy Department of Information Technology, KL University, Vijayawada, AP, India. 1 2 3 4 khaleel420@ 5 Abstract As there is an enormous growth in the web in terms of web sites, the size of web usage data is also increasing gradually. But this web usage data plays a vital role in the effective management of web sites. This web usage data is stored in a file called weblog by the web server. In order to discover the knowledge, required for improving the performance of websites, we need to apply the best preprocessing methodology on the server weblog file. Data preprocessing is a phase which automatically identifies the meaningful patterns and user behavior. So far analyzing the weblog data has been a challenging task in the area of web usage mining. In this paper we propose an effective and enhanced data preprocessing methodology which produces an efficient usage patterns and reduces the size of weblog down to 75-80% of its initial size. The experimental results are also shown in the following chapters. Keywords: Web usage mining, Preprocessing, weblog. 1. Introduction Web usage mining (WUM) is one of the applications of data mining techniques which discover the usage patterns from web usage data. The outcome of this web usage mining can be used for web personalization, website modification, system improvement, and marketing etc. Generally web usage mining[4] consists of 4 stages collection preprocessing discovery 4. Analysis and knowledge discovery as shown in fig1. Data Collection Data Preprocessing Pattern Discovery Analysis and .
